The Six Most Common Cloud Migration Mistakes


By: Larry Loeb|


Enterprises moving their business from a fixed data center to the cloud face a daunting cloud migration path. Not only must critical business data and processes be seamlessly transferred to a new environment, but the ways that information is used and accessed may need to be altered as well.

Cloud migration tools can help here, but there are other things to be considered. Here are six common mistakes enterprises make when transitioning to the cloud — and how you can avoid them.

1. Ignoring Business Users

One of the most common mistakes is not engaging the business users from the beginning. The entire purpose of migration should be to make the user experience better and their tasks easier to perform no matter the technology used. If users and their feedback are not involved in the process, the migration may be a functional failure even if it is a technical success.

2. Declining Data Quality

Poor data quality in a legacy system is yet another problem. Performing an as-is analysis can help determine if any essential data is missing or in bad form. Most organizations will assume the data itself is correct, but is it in the right form for migration? For example, is it stored in one file instead of multiple sub-files?

3. Disregarding Data Validation

Not performing data validation on the legacy data is another common mistake. This will focus on making sure the data to be migrated is correct in value and has not been corrupted along the cloud migration path.

4. Neglecting Employee Skills

The employees who are handling the cloud migration may not be up to the task. While they can be excellent at their current responsibilities, they may not be up on best practices regarding tools, processes, services, templates and accelerators. Migration is not a recurring process; it’s something that employees need to handle alongside their regular duties. It may be that they have never faced this kind of challenge before in their careers.

5. Constructing Data Set Dependencies

Cross-object dependencies can really mess things up, especially when it is realized late in the migration process that a new source of data is necessary for an application to function. Knowing requirements in advance, through a detailed analysis of all necessary data sources in each of the migrating applications, may mitigate this risk and general inconvenience.

6. Saving Cloud Migration for the Last Minute

The biggest mistake you can make is trying to go live with one big upload at the end of the process. This is putting all of your eggs in one basket. You hope it works; it usually doesn’t. Testing in several steps with multiple, iterative test loads before you go live will show if the analysis and methods used are going to operate smoothly.

Cloud migration can be done well as long as the process is broken up into smaller tasks. Just throwing a big copy command out there is not migration; it’s a way to lose what you already have. Careful, straightforward work is what you need to keep an important business goal on track.

Topics: , , ,